The arXiv turned twenty-two years old back in August, and we’re beginning to see the type of experimentation and rule-breaking you would associate with a young adult. I discovered the latest example, the Selected Papers Network (SP.net), while browsing MathOverflow. I will probably regret using the phrase, but this project is a “Web 2.0 mashup” of existing social networks and the arXiv preprint server. No, SPNetwork doesn’t promise preprints on your toast, but it should offer a useful aggregation service for informal comments on journal articles. Also, hashtags.
In his paper Open peer review by a selected-papers network, UCLA Professor and SP.net creator Christopher Lee describes a selected-papers network as:
“… a network in which researchers who read, write, and review articles subscribe to each other based on common interests. Instead of reviewing a manuscript in secret for the Editor of a journal, each reviewer simply publishes his review (typically of a paper he wishes to recommend) to his SP network subscribers.”
The current implementation lashes together Google+ and the arXiv. Using the #spnetwork hashtag in a public Google+ post along with the arXiv identifier for a paper provides a hook for SP.net to latch onto your comment. It is then echoed by the SP.net system, allowing users of that website to discover yours and other comments on a paper.
On one hand, SP.net has many qualities of existing mathematical review services such as Zentralblatt MATH, Mathematical Reviews or its electronic version MathSciNet. Contrasted to these formal systems, anyone with a Google+ account may act as a reviewer on SP.net. This is not a new idea. The seemingly-defunct arxaliv.org project used the reddit system to acheive similar ends. See also the Peer Evaluation project. Or, this famous New Yorker cartoon. So why should you pay attention to SP.net? For one, it has the support of some big names in mathematics (see blog posts here, here and here), which may help it acquire the critical mass necessary for a successful social networking site.
The core sharing and recommendation features of SP.net look more like a peer-review system than a traditional literature review; and the phrase “open peer review” is right there in Prof. Lee’s paper. The second phase (of three planned) transitions SP.net from the current architecture, to an ad-hoc publishing platform. It’s hard to say exactly what this will look like when implemented, but the idea is to have the system make it easy to identify important papers, allowing editors to peruse existing reviews.